PCRaster and LUE

You want to run your PCRaster Python model on very large spatial and temporal scales? Have a look at our LUE project, a modelling framework for the development of agent-based and field-based models. It uses the asynchronous many-tasks (AMT) approach for optimal usage of available hardware resources, and resulting LUE models can be executed on laptops or on large computer cluster without changing a single line of the model code.

LUE supports the transition of PCRaster models to LUE, the majority of PCRaster operations already have LUE equivalents. We ported, for example, the PyCatch model and ran it for Africa at 3 arc-second resolution, using 12 cluster nodes containing 1152 CPU cores in total.